Levente Nov?k a ?crit : >>>- dirt on the calibration stripe or windows. Even a small dark spot on >>> the area where shading correction is done could have a big impact, as >>> it gets applied over and over for the whole scan height. >>> >> >>This is a possibility. I will check this as soon as possible : in few hours... >> >>Could you explain the calibration process with much details. Is it possible >>that >>the calibration process be different whith snapscan backend than with the >>epson >>software ? >>I mean : the position of the calibration is done under the controle of the >>firmware, or under the controle of the backend ? > > > I don't know for the Epson, but previously I had a Microtek Phantom > 636CX and there calibration was done in software, by the backend. And > yes, it happened that I had a speck of dust on the calibration stripe > (it had an impact on the scan, although the dust was small and did not > completely occlude light). >
So this was the problem : a fine particule of dust on the backlight of the tpu. But at the wrong place : ie at the calibration position (which is not at the same place with the epson software !). It's necessary to clean not only the glass, but also the backlight ! Now, the scanner work very well until the 1200 dpi. But at a higher level (2400 and 3600) it seems there is a shake of the line. Is it what it is expected ? Nevertheless, thank for all your help and advice :-)
